Use test scores to rate teachers? I object!
OK, now we have a full week of blogging and fury, a statement from Randi Weingarten, a resolution for the next DA, and what, four pieces on Edwize?
they couldn’t find experienced educators to fill all their administrator positions, so they are making experience unnecessary
So some nasty ‘centrist democrats‘ (apologies for the links, seemed necessary) go after Leo and the UFT for being opposed to rating teachers based on test scores on principle. And Leo clearly shows that they have ignored his carefully constructed methodological arguments against.
They accused EdWize having principles. Why not take a moment to thank them?
Sure, the effects of teaching are hard to measure, and there are all sorts of quantitative problems with finding them. Lots of methodological objections, many quite powerful.
But we also know that good teaching has little to do with teaching to the test. We’ve developed, through years of arbitration, bickering, grievances, and agreement, a system of evaluating teachers that, kind of, sort of, works. It looks at how the teacher teaches. And it involves people who know teaching (and not so long ago were still teaching a class) talking with people who are doing the teaching. It involves suggestions for improvement. It involves commendations for good work. And on the negative side, involves various warnings.
This value added crap removes that. On principle we should oppose a no-feedback system. On principle we should oppose a system that reduces students and teachers and schools to a single number or group of numbers. Remember? Progress Reports? Quality Reviews? Opposition to abuse of high stakes testing?
(Hmm. The principals and APs should be just as annoyed, if not more. The DoE is making them superfluous, expendable, replaceable. Any empty suit can read numbers from a chart. See, they couldn’t find experienced educators to fill all their administrator positions, so they are making experience unnecessary.)
But I have strayed.
On principle we should oppose a system that reduces students and teachers and schools to a single number or group of numbers
The objections should include:
- methodological (noted abundantly on EdWize)
- principled (in the resolution proposed for the next DA)
- questions of competence (the DoE has bungled each major piece of data it has handled recently)
- theoretical (is raising scores an end in itself; what is the effect on teaching?)
- thorny (if AP’s cheat on tests – see JFK or Susan Wagner – what pressures will our members be under?)
And we should not be embarassed to have principles.

Sounds like a silver bullet from education experts. Experts because of course they were educated. Next time I go to the dentist, maybe I should ask to drill on some one, I have had lots of cavities.
Well if the students in a teacher’s class are averaging 20-30% higher on a standardized exam than the average across all of the people who wrote it year after year, then obviously the teacher should be recognized for that. And if the teacher’s class is scoring 20-30% LOWER on the standardized exam (especially if this exam is used for post secondary entrance requirements), then obviously something drastic needs to happen, either the teacher needs to improve their teaching, or ship out.
sdgasdg: that doesn’t take into account differences in the class makeup. An honors class might well be expected to score 20-30% higher than the average across all people taking the exam, while a remedial class might well be expected to score 20-30% lower, through no skill or fault of the teacher. If groups of students perform better or worse than they should reasonably be expected to based on their overall mathematical aptitude, then that may be attributed to good or poor teaching, but how are we going to specify in advance how well each student or group is expected to do. Even two different honors classes (or two different remedial classes) with the same teacher will often have substantially different averages, not due to any differences in the teaching, but due to differences in the strengths and weaknesses and whatever else is going on in the lives of the particular students who happen to be in a class. I think it would be very hard to come up with a fair metric of teacher effectiveness based on student scores given the inherent differences between students and classes.